LIBOR, the reference rate for more than US$300 trillion of contracts globally and nearly US$200 trillion of US dollar contracts, is expected to cease after the end of 2021. In highlighting the limitations of the LIBOR reforms made following widespread allegations of LIBOR manipulation, Andrew Bailey, the chief executive of the UK Financial Conduct Authority, placed great emphasis on the inactivity of the interbank unsecured wholesale market, with the effect that panel banks were regularly providing submissions based on "expert judgement" rather than actual transactions. Accordingly, when selecting new rates, regulators have expressed a clear preference to move to near "risk-free rates", anchored in active, liquid underlying markets. In this second article in our series, we look at the risk-free rate proposed as the US dollar LIBOR replacement and the progress made towards its transition in the loan market.
US Dollar Rate: SOFR
The Federal Reserve established the Alternative Reference Rates Commitment (or ARRC) to identify an alternative to US dollar LIBOR, with the ARRC selecting the Secured Overnight Financing Rate (or SOFR) as the appropriate replacement in 2017. SOFR comprises three overnight US Treasury Repo rates. As a backward-looking rate, SOFR was first published on 3 April 2018, reflecting the transactional data from 2 April 2018. The rates are expected to be published each day at approximately 8:00 a.m. Eastern Time, together with statistics on the total dollar amount of transactions used to calculate each rate, rounded to the nearest billion. With around US$1 trillion trading every day (and rising), the US Treasury repo market is liquid and deep, thus addressing one of the main shortcomings of LIBOR highlighted by Andrew Bailey. There are however some notable differences between LIBOR and SOFR:
- LIBOR is a forward-looking rate, quoted for periods of time (for example, one month or three months). SOFR, on the other hand, is a backward-looking overnight rate, with the rate quoted after the expiration of the period to which it relates;
- LIBOR contains a built-in credit component, meaning the LIBOR rates quoted by panel banks take into account the respective banks' term rate liquidity risk and credit risk. SOFR, however, being a "risk-free rate", contains minimal credit risk; and
- LIBOR is an interbank unsecured lending rate, measuring the average rate banks with access to the wholesale, unsecured funding market could obtain funds. In comparison, SOFR is based on overnight transactions secured by US Treasury securities. It therefore represents the funding cost of secured transactions.
As a result of being both risk-free and secured, SOFR generally produces a rate lower than LIBOR. Notably however, the SOFR rate is produced after the period of time for which the quote is required, rather than before it.
SOFR is already used in the derivatives market and the expectation is that SOFR will also be used in the loan market in due course. However, progress in the loan market has been complicated by the fact SOFR is currently only available as an overnight rate. Accordingly, in April 2019, the ARRC released a white paper ("A User's Guide to SOFR") exploring different ways in which market participants in the cash markets could use SOFR. Their suggestions can be broadly summarised as follows:
- A forward-looking term rate: being the most analogous to LIBOR, a forward-looking term rate could be extrapolated from a combination of SOFR futures and overnight index swap transactions, giving a rate based on the expectation of where the interest rate should be at the end of the given term. This rate could be quoted for one, three and six month terms.
One of the objectives given to the ARRC at its conception was to ensure any alternative rates identified were consistent with the Principles for Financial Benchmarks outlined by IOSCO (being principles endorsed by the G20 leaders, forming the basis of regulation on interest rate benchmarks across the world, and are generally recognised by the market and regulators as best practice for benchmarks). Given this objective, the ARRC requires any term rate developed also be compliant with those principles. The paper notes however, given that SOFR is a relatively new rate, additional liquidity will be required in the SOFR derivatives market before a forward-looking term rate can be produced. Despite the fact the SOFR derivative market has developed quickly, the ARRC has suggested the earliest a forward-looking term rate can be expected is the fourth quarter of 2021. Accordingly, the ARRC is suggesting market participants not wait for a SOFR term rate to become available before making the transition to SOFR.
- SOFR compounded in advance: this rate would be compounded for the one, three and six month terms occurring ahead of the interest period, to match the length of the relevant interest period. For example, for a one-month interest period beginning on 1 October, the compounded rate would be calculated for the month of September.
Although this has the advantage of providing parties with certainty as to the appropriate rate (and interest due) at the start of the interest period, a major flaw is that the rates are considered stale as they reflect the state of the market during a historic period and, accordingly, parties are vulnerable to a change in interest rates.
- SOFR compounded in arrears: this rate would be compounded during the current interest period. For example, for a one-month interest period beginning on 1 June, the parties would take the daily SOFR rate and compound it each day from 1 June to 30 June.
The advantage of this approach is the interest rate reflects the exact interest rate for that relevant period rather than a historic period or a projected rate. With the derivatives market also expected to use this methodology for its fallbacks, there will be compatibility with any corresponding hedging that parties decide to put in place. One drawback of this option, however, is the final rate will not be known until the end of the interest period, which creates operational challenges for loans as borrowers typically need time to arrange for payment of interest prior to it becoming due. In addition, borrowers typically require certainty as to the precise cost of borrowing before making a request for funds.
To address these issues, one suggestion is lenders delay the receipt of interest payments for a certain number of days until after the end of the interest period. However, this option creates difficulties for lenders as they will not see any return on their investment for the period during which payment is deferred and they may have their own funding cost obligations to meet. More feasible therefore is the possible use of the "lag" mechanism (which provides for the use of SOFR from a few days earlier than the start of the interest period, and therefore ending a few days earlier to allow the borrower sufficient time to arrange payment) or, alternatively, the "lock" mechanism (which applies the SOFR rate from the first day of the interest period, but instead repeats the daily SOFR rate for the last few days of the interest period, again giving the borrower sufficient time to arrange payment). The lag mechanism has been effectively used in SONIA-linked notes issuances, whereas the majority of SOFR-referencing floating rate notes have used a lock mechanism. Interestingly however, these SOFR notes used simple averaging rather than compounding, demonstrating a further shift in approach across markets.
- Simple daily SOFR in arrears: the final suggestion is the use of a simple rate that accrues daily during the interest period, without any compounding. As with "SOFR compounded in arrears", this means the payment due is not known until the end of the interest period (which can be addressed, as noted above). More fundamentally however, the white paper notes that whilst this simple interest convention is the most straightforward approach from a transitional standpoint (since many banks' systems are already set up to accommodate it), compound interest more accurately reflects the "time value of money", which becomes an important consideration as interest rates rise. It suggests therefore that market participants may prefer the compounding approach, which, as noted in one of our earlier articles (available here), is the preferred choice for SONIA.
Preparing for the Transition – Documentation
To facilitate the transition of loans from LIBOR to SOFR, in May 2019, the ARRC developed LIBOR fallback language to help promote market consistency for the transition process.
There are two parts to the LIBOR fallback language. First is the trigger event that initiates the transition from LIBOR to a replacement rate. These are broken down into: (i) "cessation" triggers, which will be triggered when either the benchmark administrator (currently the ICE Benchmark Administration), the administrator's regulator (currently the FCA) or the central bank for the currency of LIBOR (for US dollars, the Federal Reserve) announces the administrator has or will cease to provide LIBOR permanently; and (ii) a "public statement" trigger, whereby the regulator states LIBOR is no longer "representative". Separately, the ARRC also recommended separate "opt-in" triggers should parties to a credit agreement wish to transition to SOFR ahead of the cessation of LIBOR, similar to that seen in Europe in the Loan Market Association's recommended form of loan documentation.
The second part to the LIBOR fallback language is the actual amendment process. The ARRC provided two suggestions for the syndicated loan market. The first, known as the "amendment" approach, broadly reflects the market convention that had arisen on this point, ahead of the ARRC's involvement. Here, following a trigger event, the agent and the borrower are tasked with identifying the replacement rate and spread adjustment (the ARRC have already noted a spread adjustment will be required to make SOFR more comparable to LIBOR, with both forms of fallback language incorporating this point), taking into account any recommendations by any relevant government body or relevant market convention. The "required lenders" (usually lenders representing a majority of the loans and commitments under the credit agreement) then have a five business day window to reject the proposal before it takes effect. The second approach is known as the "hardwired" approach, where, once a trigger event occurs, the waterfall of possibilities identified in the credit agreement take effect. The first stage of that waterfall is to use any available forward-looking term SOFR (plus a spread adjustment). As noted above, the availability of such a term rate is unlikely to be imminent and therefore parties will need to look to the second stage of the waterfall, which is compounded SOFR plus a spread adjustment. Should that be unavailable (which again is feasible, given the market is yet to select the most appropriate option for the loan market), the waterfall falls back to using the "amendment" approach.
Given these difficulties, it is not surprising the "amendment" approach is the most actively used approach in the market. However, to date, our experience is the ARRC suggested "amendment" approach has not gained much traction in the syndicated market. Instead, market participants continue to use the "short-form" language adopted by the market ahead of development of the ARRC's "amendment" approach. Whilst there are similarities between the market-led language and the ARRC approach (including the administrative agent and the borrower having the right to select a replacement rate subject to a negative consent right in favour of the Required Lenders), one notable difference is this market-led version rarely includes a spread adjustment. Separately, for a deal where a non-lender administrative agent has been engaged or there are only a few lenders (including many direct lending deals), we often see modifications to the market-led version so the replacement rate must be agreed by the lenders (whether all or a majority) and the borrower.
Conclusion and Next Steps
Progress has been made in the transition from LIBOR to SOFR, with ARRC considering the possible rates that could be used by the loan market and getting loan market participants ready to make the documentary changes in due course. However additional work is required—with a forward-looking term rate less of a certainty, a decision will need to be made between the options summarised above. Only once that happens can lenders look to address the operational changes that will be required before trigger events occur that warrant documentation to be amended. With regulators keen for the loan market to make the transition, the key question now is the form of the transition, as giving effect to the transition is the only option.